This study aims to propose a data-driven framework to prioritize building retrofitting strategies based on residents' demands. It combines data-driven demand extraction with a mapping mechanism to translate residents' needs into prioritized retrofit strategies, supporting objective and scalable building retrofitting decisions.
A framework integrating text mining and sentiment analysis is proposed to extract and quantify residents' demands. It calculates demand priorities by combining textual importance and dissatisfaction. An improved quality function deployment (QFD)-based model is developed to map demand categories to retrofitting approaches and rank the retrofitting approaches. Sensitivity analysis is further conducted to examine the influence of different levels of sentiment adjustment on the prioritization results.
The results show that after incorporating residents' dissatisfaction, demand priorities are concentrated on basic living conditions, particularly vertical mobility, water supply and drainage and thermal comfort. Compared with priorities based solely on textual importance, incorporating sentiment noticeably changes the ranking of several demands. Under the baseline scenario (λ = 0.5), water supply and drainage retrofit, elevator installation and waterproofing retrofit are identified as the three highest-priority strategies, accounting for 52.10% of the total weight.
The proposed approach supports policymakers and practitioners in prioritizing retrofit measures aligned with residents' needs, improving resource allocation efficiency in large-scale building retrofitting projects.
This study provides a novel data-driven framework that bridges the gap between unstructured demand information and technical retrofitting solutions. It contributes to demand-driven decision-making by enabling objective and scalable prioritization of retrofitting strategies.
